ACL issue on Catalyst 6509 with SUP1A-2GE

Hi there,

I am experiencing strange behavior on our Core L3 Switch that is running IOS release 12.1(13)E4. I configured a reflexive access-list consisting of no more than 10-15 lines and bound this access-list to a SVI.

As soon as this access-list is active on that interface, the entire InterVLAN-traffic is process-switched and the CPU utilization raises up to 80-90%.

This is the only access-list on this switch, so the problem seems not do be related to exhaustion of the tcam:

---------- *snip* ---------- core-switch#show fm summary Current global ACL merge algorithm: ODM ODM optimizations enabled Interface: Vlan317 is up ACL merge algorithm used: inbound direction: ODM outbound direction: ODM TCAM screening for features is ACTIVE outbound TCAM screening for features is ACTIVE inbound

core-switch#

---------- *snap* ----------

---------- *snip* ---------- core-switch#show tcam counts Used Free Percent Used Reserved ---- ---- ------------ -------- Labels: 4 508 0

ACL_TCAM Masks: 10 2038 0 0 Entries: 35 16349 0 0

LOU: 0 64 0 ANDOR: 1 7 12 ORAND: 0 8 0

core-switch#

---------- *snap* ----------

---------- *snip* ---------- core-switch#show int vlan317 stats Vlan317 Switching path Pkts In Chars In Pkts Out Chars Out Processor 471958343 1075914751 498524032 1275978473 Route cache 115707 30359487 60398 5746536 Distributed cache 0 0 0 0 Total 472074050 1106274238 498584430 1281725009 core-switch#

---------- *snap* ----------

Any further ideas?

Thanks in advance, Martin

read more and respond »

Posted 5 years ago by Martin Turba in Cisco Systems

Report misuse

Image for ACL issue on Catalyst 6509 with SUP1A-2GE